Osteoarthritis is a frequent human disorder with considerable socioeconomic impact. Biochemical studies have shown that osteoarthritis disease is active at a subclinical level long before a clinical diagnosis is made. No direct method of examination of articular cartilage in vivo is available, except for surgical arthroscopy which is impractical for multiple observation points. Methods We obtained MR microscopy speciments consisting of resected cartilage and bone from patients undergoing joint replacement due to OA of hip or knee joints. Regional samples of normal appearing vs osteoarthritic cartilage and were examined at 400 MHz. Measurements were obtained on corresponding histological sections for correlation. We found that 3-D, high resolution, short TE images of cartilage samples reveal excellent correlation with thin section microscopy of the same sample with respect to the measured thickness of the cartilage and cortical bone. The error in measurement of cartilage and bone thickness was found to be T2 signal loss not to partial volume effects or bulk susceptibility. This methodology makes it possible to demonstrate zones within normal cartilage, with regional contrast achieved using either MTC, Tl- or T2-weighting. In our second study, MR microscopy of osteoarthritic cartilage, we applied results from our prior study to abnormal cartilage samples from patients with knee osteoarthritis. We performed measurements on both MR microscopy images and corresponding histology sections, and found very close correlation. We examined variations of T2 measurements within the layers at different depths in cartilage, and found a pattern of increasing water content in the more superficial layers of osteo arthritic cartilage. Discussion We will apply high resolution MR imaging in patients with osteoarthritis of the knee to achieve early non-invasive diagnosis of this disorder. We will use arthroscopy as the gold standard, and will we perform longitudinal studies in patients on different treatment protocols.